Molecular basis of membrane remodeling during secretion at the plasma membrane. Secretory epithelia such as salivary glands and pancreas represent a robust model system to study various aspects of the remodeling of membranes during intracellular trafficking processes, such as regulated protein-secretion and plasma membrane homeostasis. 1) Regulated exocytosis in salivary glands. In salivary glands acinar cells, secretory proteins are packed in large granules at the trans-Golgi network (TGN) and transported to the cell periphery where they fuse with the apical plasma membrane (APM) upon stimulation of G-Protein coupled receptors, thus releasing their content into the acinar canaliculi. Concomitantly, the membranes of the secretory granules gradually integrate into the APM, thus undergoing substantial remodeling. Our aim is to elucidate the molecular machinery regulating the integration of the secretory granules with the APM. To this end, we developed an experimental system in live rodents aimed at imaging and tracking individual secretory granules. We established that upon stimulation of the beta-adrenergic receptor, the granules fuse with the APM, followed, after a short delay, by the recruitment of a complex composed of F-actin and two isoforms of non-muscle myosin II (NMIIA and NMIIB). We showed that actomyosin contractile activity regulates the integration of the granular membranes into the APM and the completion of exocytosis. Modeling of this process based on the EM ultrastructural analysis of the secretory granules and the APM, revealed that the integration is energetically unfavorable, since it is constantly opposed by a convective flow of membranes directed from the APM to the granule membranes. This process is driven by the fact the membrane tension of the APM bilayer is higher than that of the secretory granules membranes. In order to understand how the actomyosin complex drives the integration, we focused on determining how F-actin and NMII are structurally arranged on the secretory granules. To this end, we used a series of selected light microscopy techniques with higher resolution than conventional confocal and two-photon microscopy, such as Spinning Disk and Stimulated Emission Depletion Microscopy (STED). Strikingly, we discovered that both F-actin and NMII assemble around the secretory granules in distinct polyhedral cages, formed by pentagonal and hexagonal units like those described for clathrin around the endocytic vesicles, although one order of magnitude larger. This represents a novel structural organization for the actomyosin cytoskeleton, never described before. Our data suggested that the NMII cage could function to crosslink actin filaments and/or transmit the forces generated by the contractile activity to the F-actin cage, and therefore to the granules membranes. Notably, the improved temporal resolution afforded by the spinning disc microscope, enabled us to capture, for the first time, 4D datasets of the dynamics of the cages during the integration process in vivo. This revealed that F-actin and NMII are gradually recruited into stable cages which maintain constant diameter and fixed shape (assembly phase). This step is followed by 1) the rapid polymerization of F-actin directed from the actomyosin cage towards the granule membranes (compression phase), and 2) the increase of the surface density of the NMII molecules (contractile phase). Finally, both cages disassemble with NMII being released in large filaments. Our data support a novel model based on a multi-step process in which first, the actomyosin cages counteract the convective flow of the lipids from the APM and prevent compound exocytosis; second, F-actin polymerization generates forces that drive the integration, using the cage as a leverage to push the membranes toward the APM; and third, NMIIA-driven contractions generate additional forces to facilitate the integration. In addition, we further confirmed that both the F-actin and NMII cages are assembled independently. These results provide a springboard to begin investigating the biophysical basis underlying the process of membrane integration. 2) Apical plasma membrane homeostasis in salivary glands In tubular organs, lumens are formed by the APM. Their establishment and maintenance is fundamental for their physiological function. Most of the studies investigating the mechanisms regulating this process have been carried out in cell cultures or in smaller organisms, whereas little has been done in in vivo mammalian model systems. We used the salivary glands of live mice to examine the role of played by the small GTPase Cdc42 in the regulation of the homeostasis of the intercellular canaliculi, a specialized apical domain of the acinar cells, where protein and fluid secretion occur. We found that in adult mice, depletion of Cdc42 induced: 1) significant expansion of the APM, 2) increase in the length of the lateral membranes, and 3) robust stimulation of endocytic trafficking both at the basolateral and the APM, which did not affect plasma membrane identity and junctional integrity. On the other hand, Cdc42-depletion at late embryonic stages resulted in 1) complete inhibition of the post-natal development of the intercellular canaliculi, and 2) stimulation of endocytic trafficking. Overall, these results show that Cdc42 plays a fundamental role in regulating both development and maintenance of the epithelial lumen in vivo, and highlight an additional role of Cdc42 in membrane remodeling, as negative regulator of endocytic trafficking pathways. 3) Coordination between cell metabolism and membrane remodeling. Membrane remodeling is an energetically unfavorable process. During protein secretion, significant energy is required for a single exocytic event to: 1) bring together and fuse a secretory granule and the PM, as estimated for various systems and integrate the granule membranes into the APM and maintain its homeostasis. We sought to determine how cellular metabolism is linked to protein secretion. First, we confirmed that regulated exocytosis is dependent on mitochondrial metabolism. Indeed, granule fusion and integration required the activity of the mitochondrial ATP-synthase, whereas they were independent from the activity of complex I. This finding suggested that beta-adrenergic signaling altered mitochondrial metabolism in a novel fashion. Second, we discovered that in vivo there are two spatially distinct populations of mitochondria in the acinar cells: one localized at the basolateral PM, and the other dispersed throughout the cytoplasm. Third, we discovered that the onset of exocytosis triggered the depletion of the intracellular ATP pool, as revealed by the fast activation of the cytosolic energy sensor, AMP-activated protein kinase (AMPK). AMPK has been shown to upregulate mitochondrial metabolism and homeostasis, and consistently, we discovered that the cytoplasmic pool of mitochondria increased motility and size. This last feature has been associated with an increase in the mitochondrial capacity to produce ATP. Notably, we found that, the increase in size of the individual mitochondria was mediated by the PKA-dependent inhibition of the activity of the mitochondrial fission protein DRP1. These studies underscore the importance of studying the spatiotemporal regulation of mitochondrial structure and function in intact tissues in vivo and they will open new avenues in the coordination of cell metabolism and remodeling.